List your company

Plane Geometry Problem Solving on MathV (test)

84.45AccuracyQwen3-VL-7B-Instruct (GeoCode trained)
45.60655.690565.77575.8595Jan 29, 2026Feb 1, 2026Feb 5, 2026Feb 9, 2026Feb 13, 2026Feb 17, 2026Feb 21, 2026

Evaluation results

MethodLinks
Qwen3-VL-7B-Instruct (GeoCode trained)Backbone=Qwen3-VL-7B-Instruct, Training Setting=Ours (GeoCode trained)2026.0284.45
Qwen3-VL-7B-Instruct (Baseline)Backbone=Qwen3-VL-7B-Instruct, Training Setting=Baseline2026.0283.19
Gemini2.5-ProTraining Data Size=-2026.0181.3
MLLM InterpreterTraining Data Size=5.5k2026.0180.8
GLM4.1-VTraining Data Size=-2026.0180.4
Gemini2.5-FlashTraining Data Size=-2026.0179.4
Qwen2.5-VL 32BTraining Data Size=-2026.0166.8
GF ReasonerTraining Data Size=184k2026.0164.9
Claude-Opus-4.1Training Data Size=-2026.0163.1
Qwen2.5-VL-7B-Instruct (Baseline)Backbone=Qwen2.5-VL-7B-Instruct, Training Setting=Baseline2026.0260.92
Claude-Sonnet-4Training Data Size=-2026.0160.7
Qwen2.5-VL-7B-Instruct (GeoCode trained)Backbone=Qwen2.5-VL-7B-Instruct, Training Setting=Ours (GeoCode trained)2026.0257.56
G-LLaVATraining Data Size=170k2026.0156.7
GPT-4oTraining Data Size=-2026.0147.1